See what your Claude agents are doing.
Sessions, tokens and cost per engineer, streamed from Claude agents over OpenTelemetry.
- 5 min
- setup, per provider
- 90 days
- history, instantly
- Same-day
- anomaly alerts
- Read-only access
- 14-day free trial
- No credit card required
Spend vs Budget
Forecast $45,000 this month
How does StackSpend handle Claude Usage Monitoring?
StackSpend monitors Claude usage from Claude Code, Cowork, and Office agents through OpenTelemetry, tracking estimated cost by source, user, model, and session. A daily signal and anomaly detection flag when a workflow starts generating more model calls than expected.
How does it work in practice?
- 01
StackSpend ingests Claude OpenTelemetry and breaks estimated cost down by source, user, model, and session.
- 02
A daily signal surfaces usage; anomaly detection flags a workflow that starts spending unexpectedly.
- 03
Usage sits beside your Anthropic API and other AI spend in one view.
What makes this work?
Price the tools that have no billing API.
Claude Code, Cowork and Office agents emit OpenTelemetry usage straight to a StackSpend endpoint — nothing to install, no credentials stored. Usage is priced per model, user and session as a token-based estimate, and labelled as one.
How it worksEvery model you run, in one lens.
Usage by base model, project and user, in tokens and in API-equivalent value. Estimated and billed usage stay separate, so the numbers never double-count and never pretend to be your invoice.
Business plan
How it worksOne message each morning. Nobody opens a billing portal.
Team plan and above
How it worksSee this running against your own bill by tomorrow morning.
Read-only · 5 minutes per provider
Who uses this?
- Product and engineering teams that need model-level visibility before AI bills surprise them.
- Buyers consolidating OpenAI, Anthropic, Claude, Cursor, or open-model spend into one operating view.
- Teams that need alerts and forecasting, not just retrospective usage dashboards.
What does StackSpend track?
- Claude Code, Cowork, Office agent telemetry
- Estimated cost by source, user, model, session
- OpenTelemetry ingest health
- Daily signal and anomaly alerts
- Usage from connect onward
When does this use case fire?
- Internal Claude workflows become daily operations
- Agent loops increase requests per task
- Long-context processing repeats without caching
- Model routing shifts toward pricier usage
Claude product usage (Code, Cowork, Office agents) is telemetry, not an API bill, and is easy to leave unmonitored.
Agentic workflows can create far more model calls than the team expects.
Without a signal, usage growth is invisible until it is significant.
How does StackSpend do this?
No native cross-source Claude usage view is built for different jobs. Here is what StackSpend adds.
No native cross-source Claude usage view
- Claude Code, Cowork, and Office usage are separate
- No unified estimated-cost view
- No anomaly alert when a workflow spikes
- No daily signal
StackSpend
- Claude Code, Cowork, and Office in one view
- Estimated cost by source, user, model, session
- Anomaly detection per workflow
- Daily signal and history from connect
Native tools show you last month. StackSpend tells you tomorrow.
Claude Usage Monitoring starts from day one — no manual setup and no threshold tuning required.
Read-only access · Flat plans, never a % of your bill · No credit card required
What do you get when you connect?
- Setup time
- Most teams can connect and validate setup in about 5-10 minutes.
- Access model
- Read-only credentials only. StackSpend does not modify provider resources or billing settings.
- Signals
- Daily Slack or email updates, anomaly alerts, and budget tracking in one workflow.
- History and forecast
- Historical spend context plus pace-to-forecast so overruns are visible before month-end.
Claude Usage Monitoring, answered
How do I monitor Claude usage?
StackSpend ingests Claude OpenTelemetry from Claude Code, Cowork, and Office agents and tracks estimated cost by source, user, model, and session, with a daily signal and anomaly alerts when a workflow spikes.
Is Claude usage the same as the Anthropic API?
No. Claude usage monitoring covers Claude products (Code, Cowork, Office agents) via OpenTelemetry; the Anthropic provider covers API billing. Use both if you use both.
Does StackSpend backfill Claude history?
No. Claude telemetry is collected from the point you configure the StackSpend endpoint forward — it cannot pull the prior 90 days.
How do I track Claude Code usage across my engineering team?
Point Claude Code’s OpenTelemetry output at StackSpend’s ingest endpoint and every developer’s sessions start appearing with estimated cost by user, model, and session. That gives engineering leads a per-person view of Claude Code usage — who uses it, how heavily, and what it costs — without collecting screenshots or self-reports.
Can I see which Claude models and sessions drive usage?
Yes. Usage is broken down by model (Haiku, Sonnet, Opus tiers) and by session, so a long-running agent workflow or a model-tier shift stands out the day it happens, with anomaly alerts when a workflow spikes.
Should I build my own Claude usage tracker from OpenTelemetry?
Claude Code emits standard OpenTelemetry, so you can stand up your own collector and dashboards. You then own the pipeline: a collector to run, storage to size, token-to-cost mapping to keep current as models change, and alerting to build. StackSpend is that pipeline as a product — point the OTel endpoint at it and per-user cost, dashboards, and anomaly alerts are included. Free 14-day trial, plans from $29/month.
Tomorrow morning: one number, in Slack.
Connect read-only today. Claude Usage Monitoring starts from day one — no manual setup, no threshold tuning required.